You are invited to an IEOR MONDAY SEMINAR…

Optimizing Retail Assortments Based on Sequential Consumer Choice

Stephen A. Smith

J. C. Penney Professor

Leavey School of Business

Santa Clara University

Abstract:

The multinomial logit (MNL) model has been successfully used in marketing studies to predict both product choice and retailer choice for consumer package goods. For products that customers purchase less frequently, such as electronics and durable goods, conjoint analysis has recently been applied to customer responses from Internet surveys to obtain the utility estimates needed for MNL choice models. This paper develops a mathematical programming formulation for optimizing a retailer’s assortment, based on a sequential MNL model of customers’ retailer choice and product choice decisions. A newsvendor model is used to represent the retailer’s inventory management costs. It is shown that the resulting nonlinear optimization can be reduced to a mixed integer LP. This formulation generalizes previous retail assortment optimization methods by allowing heterogeneous customer utilities and incorporating a sequential MNL choice model similar to those developed in marketing. A commercially obtained data base of customer utilities for DVD players is used to illustrate the optimization methods and the results.

Biographical Sketch:

Stephen A. Smith is J.C. Penney Research Professor and Associate Director of the Retail Workbench Research and Education Center in the Leavey School of Business at Santa Clara University. Professor Smith teaches operations management and computer modeling classes and has served as Chair of the Operations and MIS Department. He is currently Associate Editor of Manufacturing and Service Operations Management and has served on the editorial boards of Operations Research, Management Science and Industrial Engineering Transactions. Professor Smith’s research focuses on supply chain management and pricing, and his publications have appeared in Operations Research, Management Science, Marketing Science, Journal of Retailing and Journal of Marketing Research, and he is currently working on a book titled Retail Supply Chain Management. He has consulted for a variety of retailers including Target, The Gap, Sears and Levi Strauss. He also served as technical advisor to three software companies that developed decision support systems for retailers and was recently granted a United States patent for a retail forecasting algorithm.

3108 Etcheverry

Monday, 25 September 2006

3:30PM-4:30PM

Come early! Refreshments will be served at 3:00PM.

Next Seminar:

October 2 : TBA